analytic study of the tadoma method

Transcription

Journal of Speech and Hearing Research, Volume 29, 332-347, September 1986
A N A L Y T I C S T U D Y OF T H E T A D O M A M E T H O D : L A N G U A G E
A B I L I T I E S OF T H R E E D E A F - B L I N D
SUBJECTS
CAROL CHOMSKY
Harvard Graduate School of Education, Cambridge, MA
This study reports on the linguistic abilities of 3 adult deaf-blind subjects. The subjects perceive spoken language through
touch, placing a hand on the face of the speaker and monitoring the speaker's articulatory motions, a method of speechreading
known as Tadoma. Two of the subjects, deaf-blind since infancy, acquired language and learned to speak through this tactile
system; the third subject has used Tadoma since becoming deaf-blind at age 7. Linguistic knowledge and productive language are
analyzed, using standardized tests and several tests constructed for this study. The subjects' language abilities prove to be
extensive, comparing favorably in many areas with hearing individuals. The results illustrate a relatively minor effect of limited
language exposure on eventual language achievement. The results also demonstrate the adequacy of the tactile sense, in these
highly trained Tadoma users, for transmitting information about spoken language sufficient to support the development of
language and learning to produce speech.
Canada is described in a recent survey article (Schultz et
al., 1984). Schultz et al. report on the use of Tadoma with
students of varying disabilities, both as. a primary means
of speech training and in conjunction with other methods
of speech and language training. The method was most
widely used from 1920 to 1960, and its use has apparently
declined since then. The survey reports that there are
some 15 to 20 deaf-blind persons in the United States
today who rely on Tadoma as their primary means of
speech communication.
There was little discussion of the method in the research literature until the late 1970s, when reports on the
speechreading abilities of experienced Tadoma users
began to appear (see above). This recent research is
motivated by an interest in developing tactile aids for
speech communication by the hearing-impaired and the
deaf-blind. The study of Tadoma is relevant to this goal in
the information it provides about the capabilities of the
tactile sense and the parameters involved in the use of
Tadoma. The adequacy of the tactile sense for processing
temporal information such as speech is clearly a question
of basic importance with regard to the feasibility and
design of such tactile aids. The degree of success that can
be attained by Tadoma users in processing speech and
developing language thus has serious implications for the
potential of tactile aids to transmit spoken language
information to deaf individuals.
A preliminary probe of the language knowledge of one
deaf-blind Tadoma subject appeared in Norton et al.
(1977). The purpose of this report is to extend the study to
additional language measures on the original subject, and
to present results of language testing with 2 additional
highly experienced deaf-blind Tadoma users.
The 3 subjects are totally deaf and blind, 2 of them
since they were l~& years old. To examine the language 1
A number of papers have recently appeared on the
Tadoma method of speechreading, a vibrotactile method
of speech perception used by deaf-blind subjects (Norton
et al., 1977; Reed, Doherty, Braida, & Durlach, 1982;
Reed, Durlach, & Braida, 1982; Reed, Durlach, Braida, &
Schultz, 1982; Reed et al., 1985; Reed, Rubin, Braida, &
Durlach, 1978; Snyder, Clements, Reed, Durlach, &
Braida, 1982). This method of speechreading has been
used in training deaf and deaf-blind individuals for both
receiving and producing speech, and for developing a
knowledge of language (Alcorn, 1932; Gruver, 1955; Van
Adestine, 1932; Vivian, 1966).
In the Tadoma method, the person receiving speech
places a hand on the face and neck of the speaker and
monitors the articulatory motions associated with normal
speech production. In the typical hand placement, the
thumb rests lightly on the talker's lips and the fingers
spread out over the face and neck (Vivian, 1966). For the
deaf-blind Tadoma speechreader, there is no auditory or
visual input. Speech perception is achieved through the
tactile sense alone. One advantage of using Tadoma for
speechreading is that Tadoma users can receive speech
from virtually any speaker, and thus are not limited to
communication with specially trained individuals with
whom they share a manual system of communication
(Reed et al., 1985; Schultz, Norton, Conway-Fithian, &
Reed, 1984).
Training in the skilled use of Tadoma for receiving and
producing speech may extend over years of intensive,
individual instruction. Students first receive training in
speech reception, followed by training in speech production through imitating a teacher's articulatory motions
(Schultz et al., 1984). In learning to produce speech, the
student monitors the teacher's articulation by placing a
hand on the teacher's face and neck, and then attempts to
match the articulation while placing a hand on his or her
own face.
The extent of the use of Tadoma in schools for the
hearing-impaired and deaf-blind in the United States and
© 1986, American Speech-Language-Hearing Association
1Throughout the body of this paper, the term language is used
to refer to English. This is a study of the English of our subjects,
and even though American Sign Language is mentioned, we did
332
0022-4685/86/2903-0332501.00/0
CrlOMSKY: Tadoma Method
of these Tadoma users, we administered several standardized verbal intelligence measures, a syntax test in use
with deaf populations, and a number of special-purpose
linguistic tests constructed for this study. In addition we
analyzed samples of their oral and written language. The
study is an exploratory one, and our purpose was to
sample a wide range of diverse abilities, rather than
attempt an exhaustive and systematic account in any one
area. We thus have included in the testing an examination
of vocabulary, and a range of syntactic, semantic, and
prosodic features of language.
METHOD
SUBJECTS
Subject LD
LD, age 55, has b e e n totally deaf and blind since age 19
months, following a case of spinal meningitis. His development was normal until that time, but he emerged from
a 9-week coma having lost both sight and hearing. His use
of language ceased, and he received no language training
for almost 4 years. At age 5:4 (years:months) he entered
the Perkins School for the Blind in Watertown, MA, and
his language training through Tadoma was begun. Records of his early school years, reported by Stenquist
(1974), provide details of his progress. Within 8 months
he had 40 expressive words, and by age 7 he had 410
words and was combining 3 words into sentences. His
schooling continued at the Perkins School until age 20, at
which time his Stanford Achievement Test scores show
an average Grade Equivalent of 6.6, with Grade Equivalents of 7.6 in Language Usage, 9.0 in Spelling, and 7.0 in
Word Meaning. LD lives today in his own home, is
married, and holds a factory job. Tadoma is his primary
means of communication, aided occasionally by tactile
finger-spelling and sign which he learned as an adult. His
oral language and Tadoma speechreading abilities are
sufficient for him to engage in fluent conversation with
untrained hearing individuals. His literacy skills enable
him to read Braille and type his own letters.
The results of audiometric testing indicate no puretone response in the right ear, while minimal low frequency response (probably vibrotactile) was observed in
the left ear. L D demonstrates no ability to discriminate or
identify speech sounds auditorily. He has attempted to
use a variety of hearing aids with no success. Results of
visual testing indicate no measurable visual acuity.
Subject RB
RB, age 49, has been deaf and blind since 20 months of
age, as a result of spinal meningitis. His development was
not analyze the subjects' competence in this area.
333
normal until that time. When he was 2~6 he entered St.
Mary's School for the D e a f in Buffalo, NY, where he
remained until age 17. He was trained in Tadoma for
speaking and receiving speech, and in the use of sign
language, signed into his hand. After graduation from St.
Mary's, he went on to study electronics at the Burgard
Vocational High School. Today he works as an electronic
technician, is a licensed ham radio operator, and has
learned computer programming. 2 RB is able to read
Braille and type his own letters. He is fluent in American
Sign Language, and he often uses tactile sign or fingerspelling in communicating with people who command
these manual systems. He reports that his communication
with others is achieved about half the time using sign
language, and half the time using Tadoma. His Tadoma
proficiency and oral language are adequate for him to
engage in conversation with untrained hearing speakers.
Results of pure-tone testing indicate minimal low frequency response (probably vibrotactile) in the right ear,
with no response in the left ear. RB demonstrates no
ability to discriminate or identify speech sounds auditorily and has never used a hearing aid. Visual tests reveal
no measurable visual acuity.
Subject JC
JC, age 54, developed normaly until age 7, when she
lost both sight and hearing as a result of spinal meningitis.
She was subsequently trained in Tadoma at the Arizona
State School and the California School for the Blind. She
attended the University of the Pacific where she obtained
a B.A. in Sociology. Today JC works for a State Department of Rehabilitation as the state-wide consultant for
deaf-blind persons. JC reads Braille and is able to type
her own correspondence and original short stories.
Tadoma is JC's primary means of communication, and she
is proficient enough to engage in fluent conversation with
untrained hearing individuals. Her spoken language is
sufficient to enable her to lecture at conferences which
she attends in connection with her employment. She has
not received training in sign language, but is skilled in
tactile reception of fingerspelling.
JC has no measurable hearing or sight. She has no
response to pure tones in either ear across the audiomettic frequencies. She demonstrates no awareness of
speech sounds and has never used a hearing aid. Similarly, tests of vision reveal no measurable visual acuity.
DESCRIPTION OF TESTS
Verbal Subtests from WAIS, WISC-R and
Stanford-Binet
Verbal subtests from three intelligence scales in use
with the hearing population were selected to provide
2RB uses Morse code to talk on his ham radio, and reads the
computer screen by tactile Morse code output.
334 Journal of Speech and Hearing Research
some standardized measures of the subjects' verbal abilities. Performance was measured on tests of vocabulary,
differences between abstract words, proverbs, and the
like. These tests were all administered orally, with a copy
in Braille available for reference in case of doubtful
perception. The subjects often referred to the Braille
copy to be certain they were perceiving the words and
questions accurately.
Two subtests were administered from the Wechsler
Adult Intelligence Scale (WAIS, 1955). The WAIS Vocabulary subtest, which requires the subject to provide word
definitions, has a maximum raw score of 80 and an
average scaled score in the range of 8 to 12 (with a
maximum of 19). The WAIS Similarities subtest, which
requires a description of how two items are alike, has a
maximum raw score of 26 and an average scaled score
identical to that for the Vocabulary subtest. The Vocabulary subtest from the Wechsler Intelligence Scale for
Children (WlSC-R, 1974) was administered as well to
provide a fuller picture of vocabulary knowledge, even
though the scaled score is applicable only to chronological age 16. The maximum raw score on this test is 64.
Finally, four subtests of the Stanford-Binet Intelligence
Scale (1960) were administered. These included (a) Differences between abstract words (which requires a subject to define differences between three pairs of words,
e.g., laziness and idleness); (b) Essential differences
(which requires the subject to describe the principal
difference between three pairs of words, e.g., work and
play); (c) Abstract words III (which requires definition of
five words, e.g., generosity); and (d) Proverbs (which
requires the subject to relate the meaning of several
proverbs, e.g., "All that glitters is not gold").
Test of Syntactic Abilities
The Test of Syntactic Abilities (Quigley, Steinkamp,
Power, & Jones, 1978) was administered to provide a
measure of the subjects' abilities with reference to norms
for deaf individuals. This test was developed for evaluating the English skills of deaf pupils. It contains a 120item screening test to evaluate performance on a set of
nine grammatical structures (e.g., negation, conjunction,
verb processes). The test items, which are presented in
multiple-choice format with four alternatives, were administered in Braille. Normative data on a population of
505 students ages 8 to 18 with a hearing loss greater than
20 and less than 120 dB are available. Average performance in the norming group is 68% correct.
Special-Purpose Linguistic Tests
Special-Purpose linguistic tests were designed for this
study to examine the subjects' knowledge of a range of
syntactic structures and the principles of semantic interpretation of syntactic structures. Particular prosodic features of language were also studied. The tests cover a
variety of grammatical properties within the domain of
29 332-347 September 1986
transformational-generative grammar (Chomsky, 1965,
1975), examining the subjects' interpretations of sentences that involve fairly complex and subtle grammatical
features of English. These are aspects of grammar that, for
the most part, would not have been taught to the subjects
but that they would have had to acquire independently
through experience and exposure to language use. The
features selected are ones that are commonly known to
native speakers of English. The test questions have all
been answered successfully by hearing native Englishspeakers, both high school students and a variety of adult
volunteers whom we have questioned informally. We
were interested in the degree to which the subjects have
been able to acquire these basic but complex language
forms, and the nature of the deficits, if any.
A Braille copy of the structural tests was prepared, and
the subjects read the questions and reviewed them orally
before responding. Considerable care was taken to ensure that the subjects perceived the questions accurately.
For the tests of stress and intonation, the test materials
were spoken to the subjects. They gave all of their
answers orally.
The items included in each of the Special-Purpose tests
are listed in the Appendix, and a brief description is
included here. The individual tests contain different
numbers of items and scores are reported as percentage of
items correct.
Structure. The structural tests examine the subjects'
knowledge of syntactic and semantic features of English.
Subjects are asked, for example, to report on the meaning
or acceptability of sentences that contain semantic complexities, ambiguities, or grammatical anomalies. On tests
entitled Deletions, Article Switch, Ambiguity, and Illicit
Comparison and Conjunction, subjects answer questions
about meaning of the sentences, the correctness of the
sentences, or how two sentences differ in meaning. On
Tag Questions, Contractions, and Phrase Analysis, subjects are asked to produce target linguistic forms.
For example, on the Article Switch test, the task is to
describe the difference in meaning between two sentences which differ in placement of a and the:
1. Maggie looked at the puppy at Peter's Pet Shop, but
later she decided not to buy a puppy.
2. Maggie looked at a puppy at Peter's Pet Shop, but
later she decided not to buy the puppy.
In sentence 1 Maggie saw a particular puppy at the
shop and later decided not to buy any puppy at all. In
sentence 2 Maggie saw a puppy at the shop and later
decided not to buy that particular puppy.
On the Ambiguity test, the task is to give more than one
meaning for sentences such as "The long drill was boring" and "The chicken is ready to eat."
Prosodics. These tests examine subjects' knowledge of
and ability to utilize intonation and stress cues to derive
meaning in phrases and sentences. The items consisted of
Compound Noun Stress, Contrastive Stress, and Yes/no
Question Intonation. In Compound Noun Stress, for example, subjects are asked to distinguish the meanings of
GREENhouse (special place for growing plants) and
green HOUSE (a house which is green).
CHOMSKY: Tadoma Method
In contrast to the other language tests, which were all
presented in Braille or had Braille copies available for
reference, the prosodic tests were delivered orally to the
subjects. These tests examined the subjects' ability to
perceive the prosodic features through Tadoma, as well
as their recognition of the linguistic function of the
prosodic information, if perceived. Sentences were repeated as often as necessary to ensure optimal tactile
access to the prosodic features pronounced by the
speaker.
Developmental Sentence Scoring
The Developmental Sentence Scoring (DSS) procedure of Lee's Developmental Sentence Analysis (Lee,
1974) was applied to a sample of each subject's spontaneous speech, produced during normal conversation in
the laboratory with members of the research group. The
DSS procedure analyzes 50 complete consecutive sentences spoken by a subject, scoring occurrences of pronoun usage, verb types, conjunctions, negatives, and the
like. Althdugh the DSS measure is intended to assess
developmental language disorders of young children and
is normed only to age 6:11, a DSS score on adult speech
can be informative in comparison to these early levels. An
adult score well above the scores of the 6-year-old norming group, for example, may be interpreted as evidence of
language development beyond the middle childhood
stage. The mean DSS score for the norming group at age
6:6 is 10.94, with a range from 8.11 to 13.78.
For each of the 3 subjects, the number of sentences
analyzed was less than the recommended 50. For this
reason, the scores reported should be considered a
"rather tentative DSS" (Lee, 1974, p. 163). In each case,
scores were calculated by dividing the total sentence
scores by the number of sentences contained in the
Sample.
RESULTS
This section presents the results for each of the 3
subjects, along with samples of their spoken and written
language. The test results across subjects are summarized
in Tables 1 through 3.
SUBJECT L D
Verbal Subtests f r o m WAIS, WISC-R, StanfordBinet (Table 1)
LD performed well on six of the seven standardized
tests, comparing favorably with the hearing population.
His WAIS Vocabulary scaled score is 12, at the high end
of average for hearing individuals. On the WISC-R Vocabulary, he scored 50 out of a possible 64. His definitions
were generally thorough and well-stated, for example,
335
TABLE 1. Scores for the 3 subjects on the Standardized Measures
for the Hearing Population.
Tests
LD
WAIS Vocabulary: Scaled (raw)
12 (54)
WAIS SimilarRies: Scaled (raw)
8 (6)
WISC-R Vocabulary: Raw (of 64)~, 50
Stanford-Binet:
Differences between abstract
83%
words
Essential differences
100%
Abstract words III
80%
Proverbs
0
Developmental Sentence Scoring
26.6
(DSS)b
Subjects
RB
7 (22)
13 (19)
32
0
JC
17 (76)
16 (22)
61
100%
33%
30%
0
24.97
100%
100%
100%
20.67
aWlSC-R Vocabulary scores are reported as raw scores only. A
scaled score cannot be assigned because the test is applicable
only up to age 16. bTentative DSS, based on fewer than 50
sentences.
sentence: "complete group of words written in one sentence; a judge gives a sentence."; calamity: "a great
disaster."
His WAIS Similarities score was within the average
range for hearing individuals, but at the low end of the
scale. He consistently described similarities between the
items by naming common attributes: "In what way are a
coat and a dress alike? .... Both made of cloth." "A dog and
a lion? .... Both have teeth, tails, four paws, both covered
w~th hair." Even with continued prompting, LD did not
provide category-type answers such as "They are both
clothing" or "both animals."
On the Stanford-Binet Essential Differences subtest, he
scored 100%, and on the Abstract Words subtests he
missed one item each. His responses were accurate and to
the point. For example, "What is the principal difference
between an optimist and a pessimist.
optimis t is a
" " P" : "An
person who looks at the bright side of something, and
who knows the best time's to come. A pessimist is a
person who is on the dark side of things and who thinks
nothing can be done."
TABLE 2. Percent correct response for the 3 subjects on the Test
of Syntactic Abilities.
Structures
LD
Negation~
Conjunction
Determiners
Question formation
Verb processes
Pronominalization
Relativization
Complementation
Nominalization
Total
100
100
100
100
90
100
89
94
8_._99
95
S~bjects
RB
JC
(Deaf norms) a
89
91
86
100
100
90
89
89
83
90
100
100
100
100
90
100
100
100
100
9--9
(83)
(64)
(78)
(73)
(63)
(67)
(59)
(65)
(65)
aBased on the data of Quigley, Steinkamp, Power, and Jones
(1978).
29 332-347
TABLE 3. Percent correct response for the 3 subjects on the
Special-Purpose linguistic tests.
Tests
Structure
Deletions
Article switch
Ambiguity: Sentences
Subject phrases
Illicit comparison
Illicit conjunction
Tag questions
Contractions
Phrase analysis
Prosodics
Compound noun stress
Contrastive stress: Pronoun reference
Focus of negative
Yes/no question intonation
LD
Subjects
RB
JC
50
0
50
40
100
83
17~
50~
100
63
75
43
"20
80
0
0
"27
100
i00
100
100
100
i00
I00
100
100
100
67"
0
0
0
75
0
0
0
100
100
100
0
qn these cases, less than the full test was administered. The
percentage listed is based on a count of correct responses with
respect to the number of items actually given. See text for details.
LD performed poorly on the Proverbs test. On this
measure, he was able to give only literal restatements of
the proverbs. In no case was he able to refer to the
generality or larger truth embodied in a proverb, and we
conclude that he does not understand the special character of proverbs.
Test of Syntactic Abilities (Table 2)
LD's overall score of 95% correct on this test places him
well above the norm of 68% for deaf subjects. His answers
were 100% correct on five of the structures: Negation,
Conjunction, Determiners, Question Formation, and
Pronomina!ization. He missed one item on Verb Processes and Complementation and two items each on
Relativization and Nominalization.
Special-Purpose Linguistic Tests (Table 3)
Structure. LD's judgments on the structural tests were
mixed. On the Deletions test, which required him to
identify missing information, he answered correctly for
the sentences that follow the general rules of English,
and incorrectly on the exceptions. For example, in answer
to "Who is supposed to wash the dishes?" in the two
sentences
1. John told Susan to wash the dishes.
2. John promised Susan to wash the dishes.
he answered "Susan" to both, instead of "John" in sentence 2. The general rule requires Susan in such constructions, and sentence 2 with the verb promise is an
exception.
He performed poorly on all four pairs of sentences on
the Article Switch test. He either did not recognize a
meaning difference in the pair, or if he thought the two
September 1986
sentences were different, described the meanings inaccurately.
On the Ambiguity (Sentences) test, he easily detected
lexical ambiguity, describing meaning differences accurately. He was successful in detecting about half of the
structural ambiguities, detecting deep and surface structure ambiguity about equally. He paraphrased the structural ambiguities that he detected quite well. For example, for They are moving sidewalks he said, "Means two
things. People are moving sidewalks, or it could mean
they are conveyor sidewalks. The sidewalks are moving."
On the Ambiguity (Subject phrases) test, LD initially
filled in only is or are. When asked if the other verb was
possible, he answered yes and reported the two meanings
for two of the five sentences.
He achieved a perfect score on the Illicit Comparison
sentences, easily and accurately describing what was the
matter with each one. For example, he rejected "This
math problem is not as hard as that rock" because "A rock
is hard to knock, to touch (he rapped on the table). Math
is hard in the head."
On the Illicit Conjunction test he also achieved a high
score. He distinguished the acceptable sentences from
the unacceptable ones, accepting the good ones immediately, and pausing at some length over the unacceptable
ones. These latter were clearly questionable to him, and
he ended up accepting some and rejecting others.
Performance was poor on Tag Questions. LD was able
to give the correct tag for only the first sentence: John is
an engineer: "isn't he?" On subsequent sentences, he resorted to a generalized tag such as "Is that so?" and was
unable to produce the syntactically accurate form. The
test was discontinued after six sentences.
Only half of the items were answered correctly on the
Contractions test. It is of interest that these were the
items that have only one possible expansion: won't,
should're, they're. The ones that he missed, you'd and
two instances of what's, are ambiguous and depend on
sentence context. You'd, for example, may derive from
either you would or you had and what's may derive from
what is, what does or what has. There was no ambiguity,
of course, in the context of the test sentences.
On the Phrase Analysis measure, LD correctly filled in
is or are in all the sentences, showing an accurate perception of the internal structure of the subject phrase.
Prosodics. LD did well on only one prosodic test,
Compound Noun Stress. On the Compound Noun Stress
test, LD gave correct meanings for hot DOG and HOT
dog, green HOUSE and GREENhouse. Black BOARD and
BLACKboard were indeterminate. LD was tested on only
these three word-pairs. His success on the first two
examples shows an ability to perceive and interpret stress
appropriately at the level of this test, where stress functions to distinguish word and phrase meaning.
LD did poorly on both Contrastive Stress tests, where
stress is used for emphasis or contrastive purposes. On
the Contrastive Stress: Pronoun Reference measure, he
did not perceive the stress differences that we pronounced in the sentences and reported that they sounded
the same to him. On the Contrastive Stress: Focus of
Negation measure, he did perceive the stress differences
as pronounced, but reported no resulting difference in
meaning.
On the Yes/no Question Intonation test, LD did not
perceive the intonation differences between the questions and the statements. He was able to correctly identify
the intonation as rising or failing about half the time, no
better than chance. When it was explained to LD that
rising intonation signals a question, he answered "That's
news to me." It is of interest, however, that in conversation LD responds entirely appropriately to such "questions." Note the following exchanges in a conversation
with LD:
You live in Kansas?
LD: Yes.
You cross the border?
LD: Yes.
Someone gives you a
ride?
LD: Yes, a friend from work.
In a taxi? By car?
LD: By car.
LD always recognizes that such constructions require
answers when they occur in context in conversation.
Intonation, which he does not perceive, is apparently a
superfluous cue under such circumstances.
D e v e l o p m e n t a l S e n t e n c e Scoring ( T a b l e 1)
The DSS procedure was applied to a corpus of 30
sentences spoken by LD in conversation with members
of the research group. LD's DSS score, on the basis of the
30 sentences, is 26.6, well above the 10.94 mean for age
6:6.
Here is a portion of LD's oral language sample:
Oh, one time one of my friends took me to a huge trucking
garage where he works. This trucking garage repairs transportation trailer trucks and trailer cabs. You know how high
they are. Well, I stayed at the garage for more than an hour
and a half or two hours and I saw all the giant mechanical
equipment there is. And I saw the small equipment for
testing and cleaning out carburetors. And I was taken to a
place where trucks were smashed up in an accident. And I
saw one cab flattened down to about a foot high. The cab-you know how big the cab is--but it was squashed down
about one foot. And I was amazed to see the trucks that got
smashed up in an accident. And my mechanics friend told me
that the driver who got out of that cab that was squashed
down by accident, got out by [?] escape. He came out alive.
He was not killed but he was very badly injured.
As can be seen in this sample, LD's spoken language is
of high quality, comparable to that of individuals with
normal hearing. His vocabulary is mature, for example,
mechanical equipment, carburetor, repairs, injured. H e
uses complex sentence structure, for example, two levels
of subordination: I was amazed (main verb), to see the
trucks (embedding 1), that got smashed up (embedding
2), and several passives: was taken, was squashed down,
was not killed. This sample is typical of his speech, which
is noteworthy for its fluency, naturalness and low incidence of error.
Note also LD's appropriate use of the verb see: I saw all
the giant mechanical equipment, I saw one c a b . . . , I was
amazed to see the trucks . . . . This use of sight verbs is
337
typical of LD's productions, Throughout his conversations, there is frequent and appropriate use of visua!
terminology. For example, elsewhere he has commented:
It's a beautiful place, I like to see the snow come and go.
Further, his definitions of sight verbs like gaze, fade, and
dazzle are exact. It is of interest that the accurate knowledge and use of sight terms has been noted for hearing
blind subjects, whose access to language is not limited as
in the case of a person who is also deaf (Landau &
Gleitman, 1985, pp. 94-97). The linguistic and cognitive
mastery of such sight vocabulary is all the more dramatic
when it develops in the absence of both sight and hearing, as in LD's case.
Written Language
A sample of LD's writing is included as another example of his language production, and to illustrate his level
of literacy. This is an excerpt from a letter typed by LD to
a member of the MIT research group.
Since my next trip will be in the summer, I hope we can find
some free time to go surfboarding to see if I can handle the
surfboard easier than waterskis, then maybe try the skis later.
Also I want to spend more time examining the train engine,
with some old workclothes on, and I hope you can find a man
who knows about all the many valves, and devices on the
engine, so he can really explain them to me.
The writing is comparable to the writing of hearing
individuals. Sentence structure is complex, including
four levels of subordination: I hope (main verb), we can
f i n d . . , time (embedding 1), to go surfboarding (embedding 2), to see (embedding 3), if I can h a n d l e . . . (embedding 4). Sentence length averages 40.5 words, indicating
a good command of the written language. There are no
errors in grammar, spelling, or punctuation. This sample
is typical of LD's productions.
In summary, LD has an excellent command of English.
His spoken and written language are fluent, mature, and
largely error-free, comparable to the speech and writing
of literate, hearing individuals. His vocabulary compares
favorably with norms for the hearing, and his syntax is
above norms for the deaf. The tests show above average
or average performance on all but one of the standardized
tests for the hearing population, with a lack of understanding of the special character of Proverbs. He scores
well aboVe norms for the deaf on a syntax test for deaf
subjects. On the Special-Purpose linguistic tests he pe rforms well on about half of the structural tests. Specifically, he succeeds with Illicit Comparison and Conjunction, deletions which follow the general rules of English,
and some contractions. He recognizes lexical ambiguity
more readily than structural ambiguity. He is unable to
provide tags for tag questions, to interpret the semantic
effect of Article Switch, and to fill in deletions that are
exceptions to general rules. The prosodic features of
language present difficulty for him. He is unable to
perceive the intonation pattern that signals yes/no questions. His perception of stress differences is variable. He
both perceives and interprets stress differences in compound nouns, but with contrastive stress shows variable
perception and no knowledge of effect on sentence interpretation.
SUBJECT RB
Verbal Subtests f r o m WAIS, WISC-R, StanfordBinet (Table 1)
RB's best performance on these tests was on the WAIS
Similarities, where his score is above average for hearing
individuals. His answers were direct and accurate: In
what way are a dog and a lion alike? "Both animals." A
table and a chair? "Both pieces of furniture." He readily
answered b y naming the category to which both items
belong for eight of the pairs.
On the WAIS Vocabulary subtest his scaled score of 7 is
just below the average range for hearing individuals. His
WISC raw score is 32 out of a possible 64. His responses
to the words that he knew were well-stated, for example,
fabric: "piece of material like cloth, made of cotton, silk";
repair: "to fix, to get things into shape."
On the Stanford~Binet Differences between Abstract
Words and Essential Differences, the only difference he
stated was between work and play: "Work is doing the
duty to do the job. In play, you don't do the duty, you're
just having fun." In all other cases, he did not know the
meaning o f o n e or both word s , and merely defined the
individual words he knew. Of the Abstract Words III, he
gave a full definition only for independent: "do anything
you please without anybody stopping you."
For Proverbs, RB was unable to give any generalizations, and merely gave literal restatements. We provided
considerable prompting and explanations of the special
character of proverbs, and modeled the answers we were
seeking, but RB persisted with literal interpretations
only.
RB's overall score on this test was 90%, well above the
norm for deaf individuals. He scored 100% on both
Question Formation and Verb Processes. He missed one
item each on Negation, Conjunction, and Pronominalization, two items each on Determiners, Relativization, and
Complementation, and three items on Nominalization.
Structure. RB's judgments on the structural tests were
mixed. On the Deletions measure, he filled in ~ e missing
information correctly for the regular constructions, and
incorrectly for most of the exceptions. The one exception
with which he was Successful was Mary is easy to see. He
29 332-347
September 1986
answered correctly for a variety of adjectives in this
sentence frame:
Mary is anxious to see. Who wants to see? "Mary"
Mary is hard to see. Who is having trouble seeing?
"The person who is trying to see Mary."
He did well on the Article Switch test, stating the
meaning difference accurately for three of the four sentence pairs. This test performance is of particular interest
because RB often omits the article a when speaking. He
uses it more regularly in his writing. He clearly understands its function in these test sentences.
On the Ambiguities (Sentences) test, RB easily detected the lexical ambiguities and explained all of them
well. He had trouble with the structural ambiguities,
detecting only two of the deep structure ambiguities,
"Flying planes can be dangerous" and " T h e chicken is
ready t o bat", and none of the surface structure ones.
On the test of Ambiguity (Subject phrases), RB initially
filled in only is or are. When asked if the other verb was
possible, he answered yes and gave the two meanings in
only one case.
He did well on the Illicit Comparison measure, explaining what was wrong with all but one of the sentences
very accurately. For example, on hearing The movie was
longer than her hair, he laughedand said, "No good. The
movie gives you the length of time. Girl's hair is a
measurement.' Difference between time and measurement."
The results on the Illicit Conjunction test were indeterminate. RB interpreted the task as one of judging if the
two events could occur together, rather than attending to
the conjoined sentence and making a judgment about its
form. His judgments about the two events were sensible
and carefully made, but not related to the point at issue
here.
He performed poorly on Tag Questions, failing to
supply any tags correctly. The task was discontinued after
four sentences, because it appeared pointless and RB lost
interest very quickly.
On the Contractions test, RB expanded approximately
one quarter of them correctly, including some whose form
varies with sentence context such as he'd. He expanded it
correctly to he had in the sentence "I knew he'd finished
his work by 5 o'clock".
On the Phrase Analysis measure, RB correctly filled in
is or are for all the sentences, assigning the correct
internal structure to all the subject phrases.
Prosodics. RB did well on only one prosodic test,
Compound Noun Stress. On the Compound Noun Stress
test, RB responded correctly to three of the four pairs. He
perceived the stress differences and correctly differentiated the meanings o{ hot DOG and HOT dog, green
HOUSE and GREENhouse, and white HOUSE and
WHITE house. At this level 0fword and phrase meanings,
he is successful in using stress cues to signal linguistic
distinctions.
He failed to make any of the relevant distinctions on
the Contrastive Stress: Pronoun Reference measure, and
it was difficult to determine whether he actually perceived the stress difference in all cases.
CHOMSK¥: Tadoma Method
On the Contrastive Stress: Focus of Negation test, he
did perceive the stxess placement on different words in
the sentence. However, he recognized no associated
difference in meaning or implication. After the function of
stress in such cases was explained to him with examples,
RB did understand and subsequently gave three correct
answers on additional sentences.
He did not perceive the intonation differences on the
Yes/no Question Intonation test. To test his pitch perception we tried singing high and low notes, and he was able
to detect large differences in pitch. The intonation
changes in the sentences, however, were apparently not
large enough for him to perceive.
D e v e l o p m e n t a l S e n t e n c e Scoring (Table 1)
The DSS procedure was applied to a corpus of 29
sentences, spoken by RB in conversation with the research group. RB's DSS score on the basis of the reduced
corpus is 24.97, well above the 10.94 mean for age 6:6.
Here is a portion of RB's oral language sample:
When I am ready to go back to work, I am thinking to take
retirement because my job is 15 miles away from home, and
I do not have good transportation. My father has to take me to
work and he is going to be 76 years old. And if I should be
working another 18 years till retirement, will he be in perfect
health for another 18 years?
This sample is typical of RB's productions. His sentence structure shows frequent use of subordination, for
example, I am thinking (main verb), to take (embedding
1), because my job i s . . . (embedding 2). His vocabulary is
mature, for example, transportation, retirement. His
speech nevertheless contains deviations from idiomatic
usage, for example, I am thinking to take retirement, in
place of 'I am thinking of retiring," and if I should be
working another 18 years in place of the colloquial "if I
work another 18 years" or perhaps the somewhat more
formal "if I should work another 18 years." This last
example reflects the common difficulty with verb tense
that many deaf speakers experience, a frequent problem
for RB. He also is sporadic in his use of the third-person
singular marker -s on verbs. Other speech samples contain examples of both the presence of-s and its absence,
for example, when I type it come out in Morse Code, my
brother goes to Montreal, maybe he knows something.
Another occasional error in RB's speech is omission of the
article a. Some examples are: so I get bit rusty, I have
IBM-PC computer, Montreal is nice place, I was going to
bring camera . . . . Again, he reflects errors common to
many deaf speakers.
339
Friday morning, we had some programs at the auditorium at
Perkins School. I had a surprise when they called my name to
give a talk about M.I.T. on the stage. I told about the different
tests like Tadoma methods and some words and etc. After the
programs, we had picnic and games in the play ground. I met
many more people, and talked with them. At 4:30, C. took J.
and me to the air port. We had supper there. E. met me in
Buffalo, and I am glad to see family and friends again.
Today I was reading Popular Mechanics and Consumer
Report and slept for a while. Tonight, I experimented with
the buzzer. This buzzer will be experiment for the electric
braille writer. I hope the braille typewriter will be here
tomorrow.
In his writing, RB uses the article a with more consistency than in his spoken language: I had a surprise, to
give a talk, for a while, although occasional omissions
occur: we had picnic, will be experiment. His writing is
for the most part grammatical. Sentences tend to be brief,
with longer sentences showing the use of subordination:
I had a surprise (main verb), when they called (embedding 1), to give a talk (embedding 2). Spelling is good,
with errors in this sample limited to word-juncture conventions in compound nouns: air port, play ground.
Punctuation is appropriate. This sample is typical of RB's
written language.
In summary, RB has a good command of English. His
spoken language is mature and fluent, and his written
language is competent. Both his speech and writing
exhibit some features common to deaf speakers, such as
lack of verb tense agreement and omission of the article a.
His vocabulary is just below the average range for hearing
speakers, and his syntax is well above norms for the deaf.
He scores above average for hearing subjects on one
standardized test (WAIS Similarities), and below average
for hearing subjects on the other standardized tests. On
the Special-Purpose linguistic tests, he performs well on
half of the structural tests. Specifically, he succeeds with
Article Switch and Illicit Comparison, deletions which
follow the general rules of English, and some contractions. He recognizes lexical ambiguity more readily than
structural ambiguity. He is unable to provide tags for tag
questions, to detect Illicit Conjunction, and to fill in
deletions that are exceptions to general rules. The
prosodic features of language present difficulties for him.
He is unable to perceive the intonation pattern that
signals yes/no questions. His perception of stress differences is variable. He both perceives and interprets stress
differences appropriately in compound nouns, but with
contrastive stress shows variable perception and no prior
knowledge of effect on sentence interpretation. Considering that RB communicates with others only half the
time using speech through Tadoma (by his own report he
uses sign language about half the time), his grasp of
English is impressively solid.
Written Language
StJBJECT J C
A sample of RB's writing is included to demonstrate
another aspect of his language production. This is an
excerpt from a letter typed by RB to a m e m b e r of the MIT
research group.
JC's language performance and all responses to the
tests are extremely high level. She is unusually sophisticated linguistically and scores well above average for the
hearing population. Of course JC's language was well
established before she became deaf and blind at age 7,
but the excellence she displays today clearly results from
continued adequate and even rich exposure to language.
She has advanced not only beyond a 7-year-old's linguistic ability, but outdistances the average hearing adult.
Verbal Subtests from WAIS, WISC-R, StanfordBinet (Table I)
JC's WAIS Vocabulary and Similarities scaled scores
are well above average, 17 and 16, respectively. She
missed only two words on the WAIS Vocabulary subtest,
giving excellent definitions throughout. For example,
breakfast: "the morning meal when one has broken the
fast of the night"; matchless: "peerless; nothing is as
good as what that is; incomparable"; travesty: "a mockery, usually something ugly; something that was beautiful
made to look ugly and obscene." Her raw score on the
WISC was 61 out of 64. Her explanations in the WAIS
Similarities were quite sophisticated, for example,
dog~lion: "They are both animals. I could have said both
are quadrupedal animals."; eye~ear: "They are both used
for receiving sensory impulses. They are both senses."
All of JC's answers on the Stanford-Binet tests were
correct. Some examples are: difference between
poverty~misery: "Poverty refers to not having earthly
goods; misery refers to pain and agony."; definition of
generosity: "Noun that refers to being generous, openhanded, very giving, unselfish, very liberal."
She interpreted the Proverbs appropriately. For "We
only know the worth of the water when the well is dry",
she responded: "We don't appreciate what we have until
we no longer have it. Or, we don't think about our
blessings until we lose them." She was even able to
construct a generalization for a proverb with which she
was not familiar. For "Large oaks from little acorns
grow", she said: "Well, I'm not sure, but it could mean
that you might have to start from the bottom but you could
build up into great strength. Like Tadoma--I started with
single sounds first and built up finally into sentences."
JC scored 99% on this test, missing only one item in
Verb Processes. When asked this particular question a
second time, she answered correctly. Her near-perfect
performance indicates full command of the syntactic
structures on this test.
JC's record on the Special-Purpose linguistic tests is
elegant and simple. She achieved a perfect score on every
one of these tests, with the exception only of Yes/no
Question Intonation under Prosodics. She completed all
29 332-347 September 1986
the structural tests and all the other prosodic tests with
100% accuracy, enjoying the type of thought they engendered and stopping to discuss meanings and words
throughout the test. There is no need to discuss her scores
individually as they were all 100%, but examples of her
responses are provided.
Structure. JC's judgments on the structural tests
showed considerable depth of understanding.
On the Deletions test, she handled the consistent
examples and the exceptions with equal facility. Her
explanations were excellent. For example, to "I told him
what to eat. Who is to eat?" She responded: "Whoever the
'him' is. The boy is to eat." For "I asked him what to eat.
Who is to eat?" she said: "The person speaking. The 'I.' "
On the Article Switch test, her responses captured the
meaning differences precisely. For sentence pair 3, the
one about the puppy in the pet shop, she said "(a) The
implication is that Maggie looked at a particular puppy,
but decided not to buy any puppy at all. (b) A puppy, one
puppy. She decided against buying this particular puppy,
but she might buy another one."
On the Ambiguities (Sentences) test, she gave excellent
paraphrases of the two meanings of the sentences, handling lexical, deep structure and surface structure ambiguity all with equal ease. For example, for "Flying planes
can be dangerous," she said "A plane flying in the air can
be a dangerous object. Flying planes yourself can be
dangerous." For "I know a taller man than Bill": "Well, it
might mean--I know a man who is taller than Bill. Or, I
know a man taller than the one Bill knows who is tall."
For "Dick finally decided on the boat": "He decided to
take the boat, or he made his decision while on the boat."
On the Ambiguity (Subject phrases) test, JC recognized
the ambiguity for all the sentences. She supplied both is
and are for each one, unprompted, and explained the
meaning differences accurately.
She analyzed the nature of the problem with the Illicit
Comparison sentences in a v e r y sophisticated fashion. In
each case she stated the two senses of the adjective,
describing the difference with precision. For example,
"The movie was longer than her hair": "Not a good
comparison. Hair is long in terms of inches. A movie is
long in terms of time or hours." "This math problem is not
as hard as that rock": "Wrong comparison as before. A
math problem is difficult to work out. A rock is hard in
terms of solidity, not in terms of working out."
On the Illicit Conjunction test, she gave good explanations of why the sentences were unacceptable. For example, "Bill called John a fool and Susan up": "No, it
doesn't fit. These two don't go together. It sounds like he
called Susan the word 'up'. It should mean he called her
up on the telephone."
On the Tag Questions test, JC finished the list easily,
answering quickly and with certainty. Her response to
the final sentence was exceptional: "The one who robbed
the bank was John, ~ "
She supplied, "wasn't it?
You had me trapped there. I wasn't sure. I almost said 'is
that not so?'--'n'est-ce pas?', or in German 'nicht wahr?' ""
She expanded all Contractions readily, 100% correctly.
On the Phrase Analysis measure, she supplied is or are
correctly for all the sentences, interpreting the internal
structure of the subject phrase appropriately.
Prosodics. JC succeeded on all prosodic tests except for
Yes/no Question Intonation. On the Compound Noun
Stress test, JC finished the test pairs successfully, and we
gave her additional items on which she also was successful, for example, BLACKbird: "species of bird" vs. black
BIRD: "any bird black in color," and FRENCH teacher:
"teacher who teaches the language French" vs. French
TEACHER: "teacher herself is French." Clearly she both
perceived the stress differences and interpreted them
correctly.
She completed both the Contrastive Stress: Pronoun
Reference and Focus of Negation tests easily and correctly. She exhibited no difficulty in perceiving the stress,
and in recognizing its use for this type of linguistic
contrast and emphasis.
Her replies in the Focus of Negation test were as
follows:
JOHN didn't sell Bill the
"A car was sold to Bill but it
car.
wasn't John who sold it."
John didn't SELL Bill the "John may have loaned Bill
car.
the car."
John didn't sell BILL the "John sold a car all right, but
car.
not to Bill."
John didn't sell Bill the
"John sold Bill something,
CAR.
but not a car."
The Yes/no Question Intonation test was the one test on
which JC performed poorly. She did not perceive the
rising/falling intonation differences, and classified most of
the sentences as statements. It is of interest that the
speaker took care to vary only the intonation while he
spoke the sentences, holding his body still. In discussion
afterward, JC explained, "I don't believe I can do it on
the basis of inflection. I go by the movements of the head.
In real life if someone was asking me something like that
they would say, 'You missed the first part of the movie?' "
(accompanied by drawing her body up and back, and
spoken with neutral intonation).
Developmental Sentence Scoring (Table 1)
The DSS procedure was applied to a sample of 36
sentences that JC spoke during a conversation with the
MIT research group. JC's DSS score on the basis of this
corpus is 20.67, well above the 10.94 mean for age 6:6.
A portion of JC's oral language sample is provided
below:
E: What is the story about how you and Judy got
together?
JC: Well, Judy was teaching a class at American River
College. A night class. It's on interpreting. She wanted me to
give a talk to the class, about interpreting for the deaf-blind.
Well, she drove me home with her since it was at night, a
night class. We had dinner, then started out. She has a big
camper-car, and what did it do but just politely stop in the
middle of the street at a stop sign and refused to go. So we sat
there. She tried to signal somebody, to ask them to call her
garage. Her gestures were unnoticed. But finally some young
man passing stopped. He was on his way to college, so he
341
phoned the garage for her, then he came back and waited.
Nothing happened. Finally he went on to college to let
everybody know, when we weren't there, what had happened. And then he came back again. Later a group of the
students and some of the parents came out to rescue us. We
never got to the class that night.
As can be determined by this sample, JC's spoken
language is of high quality, comparable to the speech of
hearing individuals. It is fluent, idiomatic and conversational. Note the use of the interesting phrase and what
did it do but just politely stop . . . . This sample is typical
of JC's oral language in its naturalness and freedom from
error.
Written Language
A sample of JC's writing is included to demonstrate her
writing ability. This is an excerpt from a letter typed by
JC to a member of the MIT research group.
Thank you enormously for all you did for me at MIT, and for
being such an overwhelmingly nice person. And great thanks
to N. also. You two together treated R. and me just about like
royalty. We certainty did appreciate everything you did for
US.
Now about that Abraham Lincoln robot in Disneyland. I
didn't think of it until we were on the plane headed for home,
and, when I did, I wondered if you or N. knew about it.
Someone told me that when the robot recites the Gettysburg
address the lips move so distinctively deaf persons can read
them. I asked my boss J. about this; she said it didn't seem to
her the lip-movement was all that distinctive. So I think you
are right: it has more to do with lighting and sound than with
any motion of the lips.
The writing is sophisticated and fluent. Sentence structure is complex, including several instances of two-level
deep subordination: Someone told me (main verb), the
lips move (embedding 1), deaf persons can read them
(embedding 2); she said (main verb), it didn't seem
(embedding 1), the lip-movement w a s . . . (embedding 2).
Vocabulary is advanced, for example, overwhelmingly,
distinctively. Grammar, spelling and punctuation are
error-free. This sample is typical of the writing in JC's
letters. She also writes short stories, for both children and
adults.
In summary, JC's knowledge of English is extraordinarily advanced. Indeed, her abilities exceed those of the
average hearing adult. Her spoken and written language
are mature, fluent, and error-free. She scored well above
average for the hearing adult on all the standardized tests,
and performed almost perfectly on the syntax test for the
deaf. She achieved a perfect score on all the SpecialPurpose tests, with the exception only of the Yes/no
Question Test in which she did not perceive intonation
differences. JC has the linguistic command of a highly
literate and sophisticated adult, and we may assume she
has reached her full potential in language.
DISCUSSION
This section summarizes the results for the language
areas studied: vocabulary, syntax, prosodics, and spoken
and written language.
Vocabulary
The vocabulary skills of the 3 subjects were good, and
compare favorably with hearing individuals. On standardized tests for the hearing (WAIS Vocabulary), JC scored
well above average and LD scored at the high end of the
average range; RB scored just below the average range.
Their definitions were of high quality and contained
many details. For example, LD defined diamond as "a
stone that comes out of the ground; very hard stone, for
rings and machine tools." and JC gave this definition for
espionage: "undercover work which includes spying and
destruction in enemy countries." The high level was
maintained throughout. On the WAIS Similarities test,
the performance of all 3 equaled or exceeded the hearing
standard.
Examples of complex vocabulary occurred in the spontaneous conversation of the subjects as well, in many
cases where a simple word would have done. For example, note the use of varieties, intricate, and opportunity
in LD's conversation: "They transferred me to another
department so I can do many varieties of work." "I do
many different kinds of jobs such as making intricate
wires for the tail l i g h t s . . . " "So I never got the opportunity to practice with the hearing aid."
On the other standardized tests drawn from StanfordBinet, performance for 2 of the subjects was excellent. JC
succeeded even with Proverbs, which were out of range
for the other subjects. Knowledge of the special character
of proverbs appears to require exposure or perhaps instruction beyond what has been available to LD and RB.
We do not have information about the source of JC's
knowledge, but it was clear she understood the principle
of proverb interpretation, as she was able to offer a correct
generalization even for a proverb with which she was
unfamiliar.
Syntax
The syntactic abilities of all 3 subjects were excellent
in comparison with a deaf population. Their scores on a
syntax test normed on a deaf population (Quigley et al.,
1978) were well above the norms for deaf speakers. In
addition, the Special-Purpose linguistic tests indicated
many areas of syntactic competence. In JC's case, there
was complete command of all structures tested, and a
high degree of metalinguistic skill as well. The other 2
subjects showed a good command of general syntax, with
areas of deficit limited to particular details of English.
The deficits are evident in the following areas on the
Special-Purpose linguistic tests. LD and RB, for example,
interpreted syntactic constructions according to the general rules of English and often failed to take account of
particular exceptions. An example is LD's processing of
the verb promise as a regular verb, rather than as an
exception, on the Deletions test. This processing according to general rules and lack of familiarity with exceptions
is consistent with reduced language exposure imposed by
29 332-347
September 1986
deafness and with early stages in child language development as well. A strength of the language learner is the
ability to construct implicit rules on the basis of a few
examples, and then to use these rules widely, extending
them to related constructions and new vocabulary. Specific exceptions must be learned one by one. Until each
one is learned, the language user assumes the general
rules apply. This was the case with LD and RB on the
Deletions test for words whose exceptional status was
unfamiliar to them. Their answers revealed knowledge of
the basic rules, and a lack with regard to specific details.
They simply used the regular rules too widely, LD for all
the exceptions and RB for all but one set: the adjectives
easy and hard in Mary is easy/hard to see. In this one
case, RB was able to interpret the structure with
easy~hard as an exception, recognizing that Mary is the
object, rather than the subject, of the verb see.
Another syntactic difficulty that LD and RB exhibited
was with Tag Questions. Neither one was able to supply
correct tags on this test. The form of a tag in English is
complex and constrained by the form of the sentence to
which it is added. Because of their complexity, tags are
typically learned fairly late by children and pose problems for foreigners learning English. They are an unusual
construction, peripheral to the basic structure of the
language. Other languages, for example, have only one
form for tags (c£ German nicht wahr and French n'est-ce
pas) rather than variable tags of complex form, and English itself offers the option of using the single word
right? as in You ordered the roast beef, right? Nonmastery of tags affects a limited aspect of the language,
not a basic structure.
The function of articles in the language is far more
basic to the language itself. English has both a definite
and an indefinite article that occur with high frequency,
and that interact in interesting and subtle ways as brought
out in the Article Switch test. Incomplete command of the
article system is a deeper problem than failure to accommodate particular exceptions or tag questions. What is
interesting here is that LD performed poorly on the
Article Switch test, although his use of articles in speech
and writing was, so far as we could observe, flawless. This
test uncovered a gap in his knowledge that was not
apparent from observations of his productive language.
By contrast, RB's speech and writing contain omissions of
the article a, but his knowledge of the article system is
complete enough to include the subtleties measured in
the test. These are interesting examples of the distinction
between linguistic competence and linguistic performance. Language production does not always reflect what
speakers know about their language, and spoken language may be an inaccurate indicator of underlying
knowledge. It is often possible, as in this case, to learn
more about specific areas of competence by probing
comprehension than by analyzing spoken language samples.
Ambiguity detection is another domain in which both
LD and RB exhibited reduced performance. Ambiguity
detection differs from the other tests of syntax in that it
relies more heavily on linguistic awareness than the other
tests. Subjects may not succeed in detecting an ambiguity, that is, notice a second meaning for a sentence on
their own, although they can recognize and confirm (or
reject) a second meaning when it is suggested. Our
testing examined detection ability, a metalinguistic skill.
Both LD and RB easily detected lexical ambiguity, and
had less success with structural ambiguity. This accords
with the developmental picture in children, in whom the
ability to detect structural ambiguity develops considerably later than detection of lexical ambiguity. Recognizing structural ambiguity appears to require greater
metalinguistic skill, which the 2 subjects have not
achieved.
In sum, the syntactic deficits of the 2 subjects tend to be
limited to marginal aspects of English, with the basic
syntax of the language largely in place. Syntactic knowledge exceeds that of most deaf persons for all 3 subjects,
and in JC's case, eqqals that of highly sophisticated
hearing speakers. JC's superior language skills, of course,
may well reflect the fact that her exposure to language,
before loss of sight and hearing, was considerably longer
than that of the other subjects.
Prosodics
These tests required the subjects to use suprasegmental aspects of the speech signal (stress and intonation) to
make lexical and syntactiq interpretations. Two separate
questions were under investigation here. One, could the
subjects perceive, with Tadoma, the physical differences
in stress and intonation that the examiner pronounced?
And, two, if the differences were perceived, could the
subjects use this information to make correct syntactic
interpretations ?
Recall that in all the other tests, perception was not
under examination. In the other tests attempts were made
to overcome any limitations of Tadoma perception by
providing Braille copies of the tests, and discussing the
wording of the tests with the subjects to make sure they
understood the questions. In these prosodic tests perception itself was examined, along with knowledge of the
linguistic role of the suprasegmental features.
All 3 subjects experienced difficulty with the prosodic
testsl The one test they all did well on was the Compound
Noun test. All 3 were able to distinguish compound
nouns like GREENhouse from the phrase green HOUSE,
reporting the meanings correctly. Clearly they perceived
the stress difference, and understood the linguistic function it serves in distinguishing compound nouns from
adjective-noun sequences.
JC was the only subject to succeed with the Contrastive
Stress tests. I n both Focus of Negation and Pronoun
Reference she made the correct interpretations, clearly
perceiving the variations in stress and understanding
their function. LD and RB performed poorly on both
Contrastive Stress tests. Although they reported perceiving the stress variations in the Focus of Negation sentences, they did not recognize any meaning differences
associated with the stress differences. On the Pronoun
343
Reference test LD did not perceive the stress differences,
and results for RB were indeterminate.
It is of interest to consider the linguistic distinction in
stress processes that LD and RB have and have not
mastered. As noted, they succeeded with Compound
Noun Stress and did not succeed on Contrastive Stress.
The compound noun/adjective-noun phrase distinction
with which they had no trouble is a stress-related syntactic and lexical process that is basic to the language. The
contrastive case that they failed to interpret, although
they perceived the stress, uses stress for emphasis and
contrastive purposes. It appears that the basic processes
are known, and it is the more peripheral processes (such
as emphasis and contrast) which are missing. Once again,
as was the case in syntax, LD and RB succeeded with
structures that are general and regular in the language,
and had trouble with the exceptional constructions.
Intonation posed a problem for all 3 subjects. None of
them perceived the differences in rising/falling intonation on the Yes/no Question test. This failure to perceive
intonation differences is consistent with the relatively
poor ability of humans to discriminate frequency changes
in tactile stimulation as documented by Rothenberg,
Verrillo, Zahorian, Brachman, and Bolanowski (1977).
Given that intonation was not available to be interpreted
as a cue in such constructions, it is hardly surprising that
the subjects were unable to succeed on this test.
Oral and Written Language
The oral language of the 3 subjects was fluent and
mature. In the case of JC and LD it is comparable to the
language of hearing individuals. RB's spoken language
contained some features common to deaf speakers, such
as lack of verb tense agreement and article omission, for
example.
The tentative DSS scores for all 3 subjects placed them
well above the 6-year-old level. All 3 scored above 20, in
contrast to the mean for the 6-year-old norming group of
10.94.
The written language of all 3 subjects was fluent and
grammatical. They did their own typing, and showed
mastery of the mechanisms of spelling and punctuation.
They all used subordination in their writing.
General Summary
The 3 Tadoma users have a command of English that
exceeds that of many deaf persons, and in many areas
compares favorably with hearing speakers. Their backgrounds differ, and they provide evidence in different
ways that spoken language can be learned effectively
through an unorthodox sensory route. Touch might seem
unlikely as a candidate for transmission of spoken language, but we see that it may function successfully for
learning language both from the early stages, as with LD
and RB who were deafened in infancy, and at the advanced level, as with JC who was deafened after language
was well established. Even when a deaf-blind individual
divides his/her already limited linguistic exposure between English and ASL, as in RB's case, spoken language
can still develop to a high degree.
In JC's case, we are interested in aspects of language
that she has learned since age 7. Her situation might be
viewed as less dramatic than LD's and RB's, as in her case
considerable vocabulary and a major portion of English
grammar were already known. JC might have managed
linguistically had she not continued to develop her language past age 7, but merely maintained what she knew
at that time. Preservation of the status quo Would have
permitted communication with others, and, from the
point of view of this study, provided satisfactory evidence
that speech ca n indeed be perceived and language preserved through touch alone. JC, however, did not remain
at the linguistic level o,f a %year-old. She has made
normal progress into mature language. Her language
today i~ in fact not only normal but extraordinarily advanced. As noted earlier, she has the linguistic command
of a 'highly literate and sophisticated adulL With the
exception of intonation which is not available through
Tadoma, all the linguistic details are in place.
JC is impressive not only in her high test scores, but
also in the detailed accuracy and linguistic finesse with
which she handled the questions. She has an extremely
well-developed sense of language; a high degree of
metalinguistic awareness, and an analytic ability with
language that amazed us. Her responses wer e the sort that
might have come from a graduate student in linguistics.
All of these results attest to JC's ability to progress with
language fully and normally, indeed to a level well above
averagel with recourse only to touch as a source of input.
It would seem that the tactile sense has enabled her to
reach full potential in language, with mastery of all detail
that we were able to test. Whereas with our other 2
subjects we observe Tadoma fostering language development from an initial (or very early) stage to a fairly
advanced level but one that lacks various linguistic details, in JC's case Tadoma has supported extensive elaboration. The information is clearly available, at the early
and at t h e most advanced levels, even through the illsuited tactile sense.
CONCLUSION
This study demonstrates that the skin is able to transmit
information about speech that is rich enough to permit
the development of language. With our subjects the eye
and the ear have been bypassed successfully and the
necessary information delivered to the brain through
touch alone. The 3 subjects also demonstrate that speech
can be successfully processed on-line without sight or
hearing, and that the tactile sense can suffice not only for
perception of spoken language but also for learning to
produce speech.
Further, our observations reveal a relatively minor
effect on language achievement of severe restriction on
amount and range of language input. Exposure to fan-
29 332-347 September 1986
guage is drastically reduced for the deaf-blind, whose
world is 18 inches away, arm's length. Language input is
available only when a conversational partner is literally
within reach, and from reading Braille. Nevertheless, on
the basis of even such limited linguistic evidence, for
these 3 subjects virtually normal language is established,
and the areas of deficit are few.
We note certain conditions that are present for our
subjects, and we are left to wonder which of them are
critical to the success of their endeavor. With regard to
background, our subjects are not multiply handicapped,
but only sensorially deprived. Brain function is normal so
far as we know. In all 3 cases, mental development was
normal up to the time of illness, and language was
developing normally.
With regard to training, various factors may contribute
to the success of Tadoma. First, the subjects received
many years of one-on-one training in this method from
devoted teachers. Second, the nature of the Tadoma
display (a talking face) is such that multidimensional
access to information about the speech signal (including
vibration, air flow, and lip and jaw movements) is provided. Th!rd, Tadoma combines learning to produce
speech with learning to perceive it. Finally, the use of the
hand in Tadoma may provide a significant reception
advantage over systems that employ other body sites.
Individual qualities may also play a rote, and it should
be recognized that our 3 subjects may not be typical of the
deaf-blind population as a whole. They certainly do not
represent individuals who were congenitally impaired.
Personal aptitude and characteristics such as inquisitiveness and drive may be important factors in a person's
ability to learn and use spoken language with a system
such as Tadoma. This unorthodox sensory route to language, though available to some individuals, may not be
equally accessible to all.
We simply do not know which of these factors, or what
others that we have not considered, are critical to the
success of our subjects. As observers of one of nature's
experiments, we can only examine the outcome and
speculate about conditions. What is clear, however, is that
language is established under conditions of extreme stimulus poverty. The human language faculty is clearly
adequate to the task of constructing a rich linguistic
system even under the unusual conditions of an impoverished stimulus delivered through an unlikely channel.
ACKNOWLEDGMENTS
The work reported here was carried out in the Sensory Communication Group of the Research Laboratory of Electronics at
the Massachusetts Institute of Technology. We are indebted to
the members of the research group for their close collaboration
throughout all phases of the research. In particular we are
grateful to CharloRe Reed for many discussions of the work and
for her extensive help in the preparation of this report.
This work was supported by the National Institutes of Health
(Grant No. 1 R01 NS!4092).
REFERENCES
AKMAJIAN,A., & HENY, F. (1975). An introduction tO the principles of transformational syntax. Cambridge, MA: MIT
Press.
ALCORN, S. (1932). The Tadoma method. Volta Review, 34,
195-198.
CttOMSKY, N. (1965). ASpects of the theory of syntax. Cambridge, MA: MIT Press.
CrlOMSKY, N. (1975). Reflections on language. New York: Pantheon Books.
FROMKIN, V., & RODMAN, R. (1983). An introduction to language. New York: CBS College Publishing.
GRUVER, M. (1955). The Tadoma method. Volta Review, 57,
17-19.
LANDAU,B., & GLEITMAN,L. (1985). Language and experience:
Evidence from the blind child. Cambridge, MA: Harvard
University Press.
LEE, L. L. (1974). Developmental sentence analysis: A grammat-
ical assessment procedure for speech and language clinicians.
Evanston, IL.: Northwestern University Press.
NORTON, S.J., SCHULTZ, M.C., REED, C. M., BRAIDA, L. D.,
DURLACH, N. I., RABINOWITZ,W. M., & CHOMSKY,C. (1977).
Analytic study of the Tadoma method: Background and preliminary results. Journal of Speech and Hearing Research, 20,
574-595.
QUIGLEY,S. P., STEINKAMP,M. W., POWER, D. J., & JONES, B. W.
(1978). Test of syntactic abilities. Beaverton, OR: Dormac.
REED, C. M., DOHERTY,M. J., BRAIDA,L. D., & DURLACH,N. I.
(1982). Analytic study of the Tadoma method: Further experiments with inexperienced observers. Journal of Speech and
Hearing Research, 25, 216-223.
REED, C. M., DUBLACH,N. I., & BRMDA, L. D. (1982). Research
on tactile communication of speech: A review. ASHA Mono-
graphs, 20.
REED, C. M., DURLACH,N. I., BRAIDA,L. D., & SCHULTZ, M. C.
(1982). Analytic study of the Tadoma method: Identification of
consonants and vowels by an experienced Tadoma user. Journal of Speech and Hearing Research, 25, 108--116.
REED, C. M., RABINOWITZ,W. M., DURLACH,N. I,, BRAIDA,L. D.,
CONWAY-FITHIAN,S., & SCHULTZ,M. C. (1985). Research on
the Tadoma method of speech communication. Journal of the
Acoustical Society of America, 77, 247-257.
REED, C.M., RUBIN, S.I., BRAIDA, L.D., & DURLACH, N.I.
(1978). Analytic study of the Tadoma method: Discrimination
ability of untrained observers. Journal of Speech and Hearing
Research, 21,625-637.
ROTHENBERG, M., VERRILLO,R. T., ZAHORIAN,S. A., BRACHMAN, M.L., & BOLANOWSYa, S.J., JR. (1977). Vibrotactile
frequency for encoding a speech parameter. Journal of the
Acoustical Society of America, 62, 1003-1012.
SCHULTZ, M. C., NORTON, S. J., CONWAY-FITHIAN,S., & REED,
C. M. (1984). A survey of the use of the Tadoma method in the
United States and Canada. Volta Review, 86, 282-292.
SNYDER, J. C., CLEMENTS, M. A., REED, C. M., DURLACH,N. I.,
& BRAIDA,L. D. (1982). Tactile communication of speech. I.
Comparison of Tadoma and a frequency-amplitude spectral
display in a consonant discrimination task. Journal of the
Acoustical Society of America, 71, 1249--1254.
Stanford-BiDet Intelligence Scale (1960). Boston, MA: Houghton
Mifflin.
STENQUIST, G. (1974). The story of Leonard Dowdy: Deafblindness acquired in infancy. Watertown, MA: Perkins
School for the Blind.
VAN ADESTINE,G. (1932). An evaluation of the Tadoma method.
Volta Review, 34, 199.
VIVlAN, R. (1966). The Tadoma method: A tactual approach to
speech and speechreading. Volta Review, 68, 733-737.
Wechsler Adult Intelligence Scale (1955). New York: The Psychological Corp.
Wechsler Intelligence Scale for Children--Revised (1974). New
York: The Psychological Corp,
345
Received August 12, 1985
Accepted January 27, 1986
Requests for reprints should be sent to Carol Chomsky, Larsen
Hall 205, Harvard Graduate School of Education, Cambridge,
MA 02138.
APPENDIX
SPECIAL-PURPOSE LINGUISTIC TESTS
The Special-Purpose tests adtninistered to the subjects
are presented in full here. Items in these tests were
drawn from a variety of sources in the linguistic literature.
Among the sources are Ftomkin and Rodman (1983) and
Akmajian and H e n y (1975).
An asterisk (*) preceding a sentence indicates an
ungrammatical sentence.
STRUCTURE
R e p o r t on S e n t e n c e M e a n i n g
(a) Deletions. Identify missing information.
1. Mary encouraged John to apply fo~ the job. Who
is to apply?
2. Mary was encouraged by John to apply for the
job. Who is to apply?
3. John is eager to see. Who is doing the seeing?
4. John is easy to see. Who is doing the seeing?
5. John told Susan to wash the dishes. Who is to
wash the dishes?
6. John promised Susan to wash the dishes. Who is
to wash the d i s h e s ?
7. I told him what to eat. Who is going to eat?
8. I asked him what to eat. Who is going to eat?
(b) Article switch. Describe the difference in meaning
b e t w e e n two sentences which differ in placement
of a and the.
la. I b u m p e d into a man on Maple Street, and when
I turned around to apologize, the man ran away.
b. I b u m p e d into the man On Maple Street, and
w h e n I turned around to apologize, a man ran
away.
2a. I didn't mind killing the chicken, but I didn't
enjoy eating a chicken afterwards.
b. I didn't mind killing a chicken, but i didn't enjoy
eating the chicken afterwards.
3a. Maggie looked at the p u p p y at Peter's Pet Shop,
but later she decided not to buy a puppy.
b. Maggie looked at a p u p p y at Peter's Pet Shop,
but later she d e c i d e d not to b u y the puppy.
4a. T h e police saw the robber on Main St., and shot
a man on Walnut St.
b. The police saw a robber on Main St., and shot
the man on Walnut St.
(c) Ambiguity
Sentences. Give two meanings for these sentences.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Is he really that kind?
The long drill was boring.
They fed her dog biscuits.
Leonard finally decided on the boat.
She hit the man with the glasses.
He bought the picture in her living room.
Congress passed a dangerous drug bill.
He kept the car in the garage.
They are moving sidewalks.
They are biting dogs.
Flying planes can be dangerous.
The chicken is ready to eat.
The shooting of the hunters was terrible.
I know a taller man than Bill.
Subject phrases. Fill in is or are, and give two
meanings according to ambiguous subject phrase.
1. Flying p l a n e s
dangerous.
2. Moving sidewalks
dangerous.
3. Exploding firecrackers ..... illegal.
4. Biting dogs
a nuisance.
5. Speeding cars
deadly.
Account f o r Ungrammaticality
(a) Illicit comparison. Tell what is wrong in sentences
in which two items (though both long, for example)
may not be compared.
1. *The movie was longer than her hair.
2. *This math problem is not as hard as that rock.
3. *John is as sad as the movie I saw last week.
4. *Hydrogen is lighter than the blue she painted
her room.
5. *Red velvet is softerthan her voice.
(b) Illicit conjunction. (6 sentence pairs: 3 acceptable,
3 unacceptable) Decide if two sentences may be
legitimately conjoined. I f not, explain.
1. John was looking for a hat.
John was looking for a pair of gloves.
John was looking for a hat and a pair of gloves.
2. Mary read an interesting book.
Mary read a fascinating magazine.
Mary read an interesting book and a fascinating
magazine.
3. John walked along the crowded street.
John walked down the steep steps.
John walked along the crowded street and down
the steep steps.
4. The station wagon looked like a good buy.
The station wagon looked like a truck.
*The station wagon looked like a good buy and
a truck.
5. Peter took his sweater off.
Peter took his time.
*Peter took his sweater off and his time.
6. Bill called John a fool.
Bill called Susan up.
*Bill called John a fool and Susan up.
29 332-347
September1986
Produce Structure Dependent Forms
(a) Tag questions. Place appropriate tag at end of
statement to turn it into a question.
1. John is an engineer, _ _
2. You and Bill have been here since
6 o'clock, _ _
5
3. You aren't certain of what you think, 5
4. Bill and I don't always agree, 5
5. These points, the chairman will take up
later, _ _
6. Mary shouldn't see him alone, 5
7. They could have been going, _ _
5
8. There were three men in the park, _ _ ?
9. In the park were three men,
5
10. Three men were in the park, _ _
5
11. For you to do that would be crazy, _ _
12. What I just said bothered you, _ _
5
13. I bet Mary won't leave today, 5
14. I expect John won't sing the songs, _ _
5
15. I don't expect John wilt sing the songs, _ _ ~
16. John is the one who robbed the bank, _ _
17. It was John who robbed the bank, _ _ ~
18. The one who robbed the bank was
john, . 5
(b) Contractions. Give full form of the contracted item.
1. What's he been doing all day?
2. He could've tried harder.
3. What's in that box on the table?
4. He won't do that again.
5. You'll never agree with me.
6. What's he want that book for?
7. He'd never been here before today.
8. I knew you'd b e good at this.
9. I should've said no.
10. i knew he'd finished his work by 5 o'clock.
11. I know he's been here before.
(c) Phrase analysis. Fill in is or are according to the
internal structure of the subject phrase containing
Verb -ing.
is: 1. Washing dishes _ _
2.
3.
4.
5.
are: 1.
2.
3.
4.
dull.
Raising flowers _ _
fun.
Knitting sweaters _ _
Satisfying.
Painting pictures _ _
hard.
Writing letters
interesting.
Sleeping children
beautiful.
Dancing bears
amusing.
Growling lions
frightening.
Swimming ducks _ _
pleasant.
PROSODICS
These tests examine the subject's knowledge of and
ability to utilize intonation and stress cues to meaning in
phrases and sentences.
(a) Compound noun stress. Distinguish the meaning of
compound nouns and adjective-noun sequences.
1. Look at that HOT dog~hot DOG on the front steps.
CI-IOMSKY: Tadoma Method
2. He stopped to look at the GREENhouse~green
HOUSE on the comer.
3. Who lives in the WHITE House~white HOUSE?
4. There are three BLACKboard erasers~black
BOARD erasers in that box.
(b) Contrastive stress: Pronoun reference. Identify
pronoun reference with normal and contrastive
stress.
1. Peter kicked Bill, and then I kicked 'im. Who did
I kick?
2. Peter kicked Bill, and then I kicked him. Who did
I kick?
3. Peter kicked Bill, and then 'e kicked Maw. Who
kicked Mary?
4. Peter kicked Bill, and then he kicked Maw. Who
kicked Mary?
5. Peter kicked Bill, and then 'e hit 'im. Who hit
who?
347
6. Peter kicked Bill, and then he hit him. Who hit
who?
Contrastive stress: Focus of negation. Identify difference in implication as different words are
stressed.
1. John didn't sell Bill the car.
(c) Yes~no question intonation. Differentiate statements from questions on the basis of intonation.
Statements (falling intonation)
1, They don't know the girl's last name.
2. It snowed again on Thursday.
3. The children are asleep already.
Questions (rising intonation)
1. She tore her sweater in the fight?
2. He hurt himself this morning?
3. He'll be here at nine tomorrow?

analytic study of the tadoma method

Transcription

Similar documents

Caring, Handling and Empowering PWDs by Dr. Josephine

Time: 6:00 to 8:00 PM Where: Shakey`s Pizza Parlor 12540 Valley

July - CaptionCall

What IDHHC Learned from the Consultation with the Hard of

May 1, 2013

Preparation assignment for virtual meeting

Humana members have access to exclusive pricing on the newest

EXCITING NEWS!!! - Sonotone Hearing Aid Center

SoundWaves-September-2013 - Canadian Hard of Hearing